Depth Perception Data Analysis with R
Description:
This project aims to analyze and visualize the egocentric depth perception experimental data based on given distances and judged distances using R. [Report]
Skills: R, Data Visualization
Graduate Course Name: Data Science with R
Year: 2019
Glimpse of Sketching
Glimpse of High Fidelity Prototype
Designing and Implementing Interface for FedEx: Flexible Worker
Description:
This is a group project, and the goal of this project is to design and implement a flexible worker system for FedEx employees. We used the "Think Aloud" technique to evaluate the users' interaction with the interface. Users also put their comments on the comment box to express their experience after interacting with the interface. We have used two surveys (pre-survey and post-survey) in our user study to gather the user experience and usability of the interface.
Achievement: Got the recognition of an outstanding effort and job well done on the FedEx HCI class UX/UI challenges from FedEx.
Skills: Flask, HTML, JavaScript, User Interface Design
Graduate Course Name: Advanced Human-Computer Interaction
Year: 2018
Investigation of Vergence Angle Error of an Augmented Reality Haploscope
Description:
An augmented reality haploscope is an optical display that imposes a controlled virtual image on the real world to enhance the user's view of the real world. A significant relationship among interpupillary distance (IPD), accommodation and vergence exist in the AR haploscope. This project focuses on measuring vergence angle error by manipulating four different IPDs at five different accommodation demands (2, 2.25, 2.5, 2.75, 3.0) diopters. The statistical analysis yields that smaller IPDs at longer distances contribute less error. In contrast, relatively large IPDs at near distances create more vergence error in an AR haploscope.
Skills: Perception, Mixed Design ANOVA, SPSS
Graduate Course Name: Quantitative Methods in Psychology II
Year: 2018
Optical Analysis of an Augmented Reality Haploscope
Description:
A Haploscope is an optical system that presents separate images for the left and right eye in a controlled way, and users can see the fused image through the optics. In this project, I analyzed the Haploscope's geometrical optical system step by step to find out the final optically displayed image. I used the general thin lens formula to calculate the formed image's properties at each optical stage of the AR haploscope. For calculating the virtual image size, I used angular magnification rather than linear magnification because the size of the virtual image is changing according to the distance.
Skills: Geometrical Optics
Graduate Course Name: Optics
Year: 2018
Stereoscopic Rendering of a 3D Model
Description:
In this project, I worked with the stereoscopic rendering of a 3D model using OpenGL. For implementing the side by side 3D model, I applied the principle of Asymmetric frustum perspective projection. For loading the 3D model on the screen, I used the Open Asset Import Library (Assimp). Finally, I executed the Phong lighting model for both left and right eye.
Skills: OpenGL, C++
Graduate Course Name: Computer Graphics
Year: 2017
Potential Energy Surface (PES) Fitting Using Machine Learning Techniques
Description:
This is a two-person group project. An approach to predict Methanol molecule's potential energy using machine learning algorithms, i.e., multivariate linear regression using gradient descent & normal equations, and Deep neural network approach using TensorFlow. A dataset of 25000 data was created using molecular simulation. Simulated data were used to train the potential energies of methanol, considering ten features. After comparing, we found that the deep neural network performs better with a reasonably small error between the two approaches.
Skills: python, TensorFlow, simulation
Graduate Course Name: Machine Learning
Year: 2017
Glimpse of Sketching
Glimpse of High Fidelity Prototype
Designing and Implementing Interface for Ramp Planning Dashboard for FedEx
Description:
This is a group project, and the goal of this project was to design and implement a ramp planning dashboard for FedEx so that ramp managers can plan and communicate their sorts for both inbound and outbound operations. Using this interface, the manager could also schedule how fast his employees need to process to hit a specific sort's start and end based on the planned and the actual packages. Further, the ramp manager received different alerts according to various potential issues such as late truck or aircraft or volume is higher than usual from our designed ramp planning dashboard.
Skills: HTML, JavaScript, User Interface Design
Graduate Course Name: Human-Computer Interaction
Year: 2017
To Test Different Methods of Tele-operation and Gauge Comfort Controlling of the Different Robots
Description:
This is a group project, and the purpose of this project is to study different methods of teleoperation. Three different modes of operation used: joystick control, voice control, and pre-planned navigation with three different types of robots: Spiderbot, Sphero, and Ollie. Results indicate that Spiderbot was the favorite in physical appearance and the favored method of teleoperation. On the contrary, Sphero was the least preferred in the appearance and mode of teleoperation.
Skills: Robot navigation, Data Analysis, User-study
Graduate Course Name: Human-Robot Interaction
Year: 2016